Original papers3D characterization of crop water use and the rooting system in field agronomic research
Introduction
Crop phenotyping techniques are rapidly advancing. Particularly with the widespread availability of proximal sensors, the use of drones, high-resolution satellite imagery, and developments in machine learning, artificial intelligence, and data integration and analysis tools such as crop simulation models (Zhang et al., 2020, Potgieter et al., 2021). However, so far, applications have involved characterizing the above-ground parts of crops, ignoring a “hidden half”: the rooting system and its activity. Here we developed and tested a new approach to produce 3D characterizations of soil crop water use and root activity in large field genotype (G) by environment (E) by management (M) experimentation.
So far, the opaque nature of the soil has limited the options for phenotyping work on root growth and activity to the use of laborious and costly root sampling methods. Research on root systems has included growing plants in chambers (Joshi et al., 2017), pots (Richard et al., 2015), intact soil cores (McNeill and Kolesik, 2004), lysimeters (Geetika et al., 2019), and rhizotrons (Atkinson et al., 2019). An important constraint of these systems is that they introduce difficult-to-control side effects that alter the growing conditions of roots and shoots (Tracy et al., 2020). In addition, when rooting systems are sampled in the field, usually a small fraction of the whole rooting system is sampled (Ordóñez et al., 2018). A large diversity of soil water monitoring approaches exist; however most approaches sample a small volume of the soil, i.e., around probes or around access tubes of neutron moisture meter. Moreover, these approaches become impractical, laborious, and cost-prohibitive, when hundreds of pots or field plots need to be assessed multiple times over the cycle of the crop. Here we claim that in field crops research, crop phenotyping or GxExM research, there is a need for fast, accurate, and repeatable 3D approaches to quantify crop water use and root activity under field conditions.
Recent, advances in geophysical techniques e.g., electrical resistance tomography (ERT), ground penetrating radar (GPR), and electromagnetic induction (EMI), offer the potential to infer crop water use within the soil profile at spatial resolutions down to a few meters (Martínez et al., 2021). ERT requires galvanic contact between electrical probes and soils, which causes disturbance of soil surface and the need for extensive electrical cabling (Huang et al., 2016). Moreover, uncertainty remains on the performance of GPR in electrically conductive media such as on heavy-clay soil. In comparison, EMI is contactless, fast, repeatable, generates spatial-dense datasets, and can be quickly applied across large spatial areas. EMI provides measures of soil apparent electrical conductivity (ECa, mS m−1) which is influenced by soil moisture (Zare et al., 2021). ECa data is also influenced by soil texture and salinity. Interestingly, the impact of temporally dependent factors (e.g., soil moisture) can be distinguished from those of temporally independent factors (e.g., texture) with time-lapse EMI measurements (Vereecken et al., 2014), providing an understanding of changes in soil moisture over time. ECa is a depth-weighted measurement that represents soil conductivity across a soil depth range (McNeill, 1980). Inversion modelling e.g., cumulative sensitivity and Maxwell equation, can then be applied to estimate a vertical profile of the true soil electrical conductivity (σ, mS m−1), and enable the prediction of soil moisture in space and depth (Zare et al., 2020). EMI was previously used to quantify crop water use and dynamics in root phenotyping studies with wheat (Whalley et al., 2017, Blanchy et al., 2020), and chickpeas (Huang et al., 2018). Here we apply similar approaches to untangle complex interactions in root water use and root activity in large on-farm genotype GxExM studies in sorghum.
The objectives of this study were to i) answer whether EMI coupled with inversion modelling, and satellite imagery can provide an approach for 3D characterizations of sorghum water use and root activity, and (ii) research whether combining 3D soil moisture dynamics mapping and simple crop-eco-physiological principles could be used to untangle complex GxExM interactions in field research.
Section snippets
GxExM study
An on-farm trial was conducted at Nangwee, QLD Australia (27°34′4.3″ S, 151°18′38.2″ E) during the 2020–2021 Southern Hemisphere summer growing season. The climate in the region is subtropical with an average 621 mm rainfall per annum and mean annual maximum and minimum temperature of 27.0 °C and 12.0 °C, respectively (Australian Government Bureau of Meteorology, 2021). The trial covered an area of ∼3.2 ha (82 m × 384 m) of a uniform black, self-mulching cracking clay, characterized as black
Model calibration and validation
Values of R2 were determined for ANN models developed using θv, estimates of σ, and soil depth with different sets of inversion parameters (Fig. 4). The largest R2 for calibration (0.78, Fig. 4a) was obtained with FS, S2 and λ = 0.3. Thus, this combination of parameters was chosen for the inversion of the ECa and the development of final ANN model. The spatial distribution of inverted σ for the two surveys was shown in Fig. S3.
The measured values were compared to the predicted θv at all the
Discussion
A new approach, in which EMI was coupled with inversion modelling, remote sensing data, and crop eco-physiological principles, was developed to produce 3D characterizations of soil crop water use and root activity in large on-farm G×M×E experimentation. Moreover, numerical and statistical techniques (i.e., PCA, LMM and recursive partitioning trees) were used to untangle complex GxExM interactions. Results indicate that the new approach could be used to produce 3D characterizations of crop water
Conclusion
This study shows for the first time how EMI coupled with inversion techniques, an ANN model, remote sensing data, and crop eco-physiological principles, can be used to characterize sorghum water use and root activity in 3D in a large field GxExM experimentation. Validation results indicated that the approach achieved a satisfactory prediction (LCCC = 0.81; RMSE = 0.03 cm3 cm-3) on the soil water dynamics and crop water use. Root activity in depth was different between irrigated and dryland
CRediT authorship contribution statement
Dongxue Zhao: Conceptualization, Methodology, Software, Formal analysis, Visualization, Writing – original draft. Joseph X. Eyre: Resources, Data curation, Writing – review & editing. Erin Wilkus: Data curation, Resources, Writing – review & editing. Peter de Voil: Resources. Ian Broad: Resources, Data curation. Daniel Rodriguez: Supervision, Conceptualization, Project administration, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work is funded by the Grains Research and Development Corporation (GRDC) project UOQ1906-010RTX. The authors acknowledge Dr. Yan Zhao and A/Prof. Andries Potgieter from UQ- QAAFI for their valuable comments on processing the satellite images. The authors also want to thank the collaborating farmers for their aid in the field experiment.
References (59)
- et al.
A refined method for rapidly determining the relationship between canopy NDVI and the pasture evapotranspiration coefficient
Comput. Electron. Agric.
(2018) - et al.
Field high-throughput phenotyping: the new crop breeding frontier
Trends Plant Sci.
(2014) - et al.
Uncovering the hidden half of plants using new advances in root phenotyping
Curr. Opin. Biotechnol.
(2019) - et al.
Phenology of short vegetation cycles in a Kenyan rangeland from PlanetScope and Sentinel-2
Remote Sens. Environ.
(2020) - et al.
Understanding the diversity in yield potential and stability among commercial sorghum hybrids can inform crop designs
Field Crops Res.
(2019) - et al.
Apparent soil electrical conductivity measurements in agriculture
Comput. Electron. Agric.
(2005) - et al.
Monitoring and modelling soil water dynamics using electromagnetic conductivity imaging and the ensemble Kalman filter
Geoderma
(2017) - et al.
Soil water extraction monitored per plot across a field experiment using repeated electromagnetic induction surveys
Soil Syst.
(2018) - et al.
Concurrent variability of soil moisture and apparent electrical conductivity in the proximity of olive trees
Agric. Water Manag.
(2021) - et al.
Probabilistic inversion of EM38 data for 3D soil mapping in central Iran
Geoderma Regional
(2016)
A solution for sampling position errors in maize and soybean root mass and length estimates
Eur. J. Agron.
Water extraction by grain sorghum in a sub-humid environment. II. Extraction in relation to root growth
Field Crops Res.
Comparing satellites and vegetation indices for cover crop biomass estimation
Comput. Electron. Agric.
Using NDVI for the assessment of canopy cover in agricultural crops within modelling research
Comput. Electron. Agric.
Crop improvement from phenotyping roots: highlights reveal expanding opportunities
Trends Plant Sci.
On the spatio-temporal dynamics of soil moisture at the field scale
J. Hydrol.
Two-dimensional time-lapse imaging of soil wetting and drying cycle using EM38 data across a flood irrigation cotton field
Agric. Water Manag.
Scope to map available water content using proximal sensed electromagnetic induction and gamma-ray spectrometry data
Agric. Water Manag.
High-resolution satellite imagery applications in crop phenotyping: an overview
Comput. Electron. Agric.
Mapping cation exchange capacity using a quasi-3d joint inversion of EM38 and EM31 data
Soil Tillage Res.
The relationships between seedling root screens, root growth in the field and grain yield for wheat
Plant Soil
Accounting for heterogeneity in the θ–σ relationship: Application to wheat phenotyping using EMI
Vadose Zone J.
Determining the within-field yield variability from seasonally changing soil conditions
Precis. Agric.
Neural network models for predicting flowering and physiological maturity of soybean
Trans. ASAE
Computer vision and machine learning enabled soybean root phenotyping pipeline
Plant Methods
Electromagnetic physical scale modeling
Electromagnetic Methods in Appl. Geophys.-Theory
Cited by (8)
In search of pragmatic soil moisture mapping at the field scale: A review
2023, Smart Agricultural TechnologyCurrent situation and development trend of well-facilitated farmland construction in China
2023, Nongye Gongcheng Xuebao/Transactions of the Chinese Society of Agricultural EngineeringThe Costs of Soil Erosion to Crop Production in Canada between 1971 and 2015
2023, Sustainability (Switzerland)Estimating the Effects of Deficit Irrigation on Water Absorption and Utilization of Tomatoes Grown in Greenhouse with Hydrus-1D Model
2023, Sustainability (Switzerland)Organic Carbon Content in Fractions of Soils Managed for Soil Fertility Improvement in Sub-Humid Agroecosystems of Kenya
2023, Sustainability (Switzerland)